Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    vanishing/exploding gradients

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     và
    0.49
    และ
    0.47
     ಹಾಗೂ
    0.44
     ۽
    0.44
     различных
    0.44
     oraz
    0.43
    例えば
    0.43
     ແລະ
    0.43
     અને
    0.42
     những
    0.42
    POSITIVE LOGITS
     इसे
    0.39
     quickly
    0.38
     أول
    0.38
     promptly
    0.37
     تع
    0.37
     newly
    0.37
     الأولى
    0.36
    पने
    0.36
     acuity
    0.35
    చి
    0.35
    Act Density 0.006%

    No Known Activations